Using Stepwise Regression Techniques to Shortlist the Number of Antecedents of Employee Absenteeism

With the help of the extended and revised model of employee absenteeism, all possible antecedents of employee absenteeism are investigated. Data is taken from a cleaning company in the Netherlands and consists of 4334 employees. With the help of stepwise regression techniques, five antecedents prove to be of most value in explaining employee absenteeism; age, gender, location and transportation problems, job demands and work group characteristics. Absenteeism was incorporated in two ways, (1) as a ratio of absent hours compared to worked hours and (2) as the total frequency of absences during a year. For three of the five antecedents, possibilities for forms are limited. And for the two antecedents with possibilties, the signs of the of relationships differed between the two absenteeism measures, which proves that solving absenteeism is a complex problem.